Driving in Dense Traffic with Model-Free Reinforcement Learning

TL;DR

This paper presents a deep reinforcement learning approach for autonomous vehicles to navigate dense traffic by implicitly modeling interactions with other drivers, enabling successful merging and lane changes where traditional methods fail.

Contribution

The work introduces a model-free deep reinforcement learning policy that effectively handles dense traffic scenarios by learning to negotiate gaps, outperforming traditional control algorithms in simulation.

Findings

RL policy outperforms model-predictive control methods in dense traffic scenarios.

The learned policy successfully negotiates gaps for merging and lane changes.

Simulation results demonstrate improved safety and efficiency over traditional methods.

Abstract

Traditional planning and control methods could fail to find a feasible trajectory for an autonomous vehicle to execute amongst dense traffic on roads. This is because the obstacle-free volume in spacetime is very small in these scenarios for the vehicle to drive through. However, that does not mean the task is infeasible since human drivers are known to be able to drive amongst dense traffic by leveraging the cooperativeness of other drivers to open a gap. The traditional methods fail to take into account the fact that the actions taken by an agent affect the behaviour of other vehicles on the road. In this work, we rely on the ability of deep reinforcement learning to implicitly model such interactions and learn a continuous control policy over the action space of an autonomous vehicle. The application we consider requires our agent to negotiate and open a gap in the road in order to successfully merge or change lanes. Our policy learns to repeatedly probe into the target road lane while trying to find a safe spot to move in to. We compare against two model-predictive control-based algorithms and show that our policy outperforms them in simulation.